RIVER Gage Work Pushed to Improve Flood Forecasting

In September 1933 $150,000 of emergency funds was allotted to the Weather Bureau to repair and improve its river gages.  Since that time the Bureau has been engaged, in cooperation with other governmental agencies, in standardizing and perfecting the gages used in river-stage and flood forecasting, and in installing gages to determine the relation of stream flow to precipitation.

The Weather Bureau has always done its river-gaging work under a handicap. Funds had never before been available for the construction at one time of more than a few gages of a substantial and modern type. A large part of the money that could be allotted for gages had to be used in maintenance, because every flood partially wrecked a comparatively large number of the structures. Of all the gages then in use, only four gave a continuous record of river stages.

The emergency allotment is, therefore, not only helpful in giving employment in several hundred widely scattered small towns, but it is furnishing to the Weather Bureau a network of river gages that will be of lasting benefit to the country.

   On June 30, 1934, there had been erected 76 staff gages, 9 of the chain and weight type, 97 of the wire-weight, and 47 continuous recorders, a total of 229. The work was finished by December 31, and all of the gages maintained by the Weather Bureau either were replaced or were thoroughly inspected and found not in need of repair.

   In addition to the 437 gages that are owned by the Weather Bureau there are 272 from which reports are furnished to the Bureau by other agencies, principally the Engineer Corps of the Army. This gives a network of 709 gage reports available for river-stage and flood forecasts. However, only 482 of the reports are made daily; 129 are furnished only during the months that may be considered to embrace the flood season, and 98 are received in times of threatened or actual flood.

   The accuracy and timeliness of the river forecasts of the Bureau have, for along time, been considered quite satisfactory by the general public. But the officials of the Bureau have always realized that the system under which the forecasts are made has an inherent disadvantage in that it has never been expressed in standardized formulas.  Each forecaster has a set of rules for the rivers in his district, but these rules must be applied in individual cases through the experience of the forecaster.  It is impossible for a forecaster to put a large part of his knowledge on paper and, when he is no longer available for this work, his successor must begin immediately to make an intensive study of the rivers in his district, and the effect on the rivers of rains of varying intensity and distribution. He must also become familiar with the relation of run-off to precipitation as it may be modified by the season of the year. The condition of the soil and numerous other things must be given consideration. Even an intensive study does not thoroughly qualify a forecaster, but actual experience must be had before he feels sure of himself.

Discharge observations or rating curves have been used to a very minor extent in Weather Bureau work. It has not been possible to employ them in any large way because they were not available.  However, since 1922, and particularly since 1927, much stress has been placed on flood protection, and, in the last few years, inland navigation and power development have progressed steadily. Hence information in regard to the quantitative flow of streams is increasing more rapidly than ever before, and a further important increase will come with the continuously recording gages put in with Public Works funds.

The Weather Bureau makes no stream-flow measurements. However, through the cooperation of the Geological Survey, rating curves will be available for strategic points on the important rivers of the country, where recording gages are situated, and the officials of the Bureau will begin a study of the application of rating curves to river-stage and flood forecasting. It is realized that these curves can be no more than an important aid; that current meteorological information will always be indispensable in river forecasting, and that if, in rehabilitating and standardizing the network of river-gage stations, the meteorological stations are neglected, no satisfactory measure of success can be attained. But it is thought the study and application of the curves will remove a reasonable amount of the personal element that now surrounds forecasting, will make it possible to refine forecasts somewhat more than at present, especially on the large rivers, and will enable a forecaster to leave for his successor formulas that are based on sound and well-understood principles.

River-stage and flood forecasting is the primary purpose of the river-gage service, but the necessity for adding another feature, the determination of the relation between stream flow and precipitation, has been growing and has increased rapidly in the last year or so.  A knowledge of this relation is necessary in making plans for power dams, irrigation projects, flood prevention and control, and farm and city water supplies. However, reliable statistics regarding the relation are too scant to be of great value, and the dry weather that has prevailed over most of the country in the last few years has shown engineers in a most positive way that sound plans for the water conservation, so necessary to agriculture and the general public, cannot be made without a definite knowledge of the volume of water streams may be expected to deliver in disastrously long periods of insufficient rain.

A knowledge of rainfall is fundamental, but this knowledge, to be of full advantage, must be extended to show what becomes of the rain after it is received by the ground, and a definite determination of the relation of stream run-off to precipitation throughout the country would prove inconceivably valuable in planning the economic life of the Nation. The climatological service of the Bureau collects precipitation data for the entire United States, and these data, in conjunction with the stream-flow rating curves prepared by the Geological Survey, will make possible the determination of this highly important relation between stream flow and precipitation.